Segment-Based Numeric Neural Network Classifier

By LeeAnne King & Sara Francis

A machine learning project that implements and experiments with segment-based numeric neural network classification. The repository contains code and notebooks for building, training, and evaluating classifiers on numeric data, with an emphasis on model architecture and interpretability.

Overview

This project explores the application of neural networks for classifying numeric data by leveraging segment-based features. The work includes data preprocessing, neural network design, model evaluation, and visualization of results. The primary goal is to understand how segment-based approaches can improve numeric classification tasks.

Skills Learned

• Data Preprocessing: Handling and preparing numeric datasets for neural network training.
• Neural Network Design: Building models in Python using libraries such as TensorFlow, Keras, or PyTorch.
• Jupyter Notebooks: Interactive experimentation, data analysis, and result visualization.
• Model Training & Evaluation: Implementing training loops, loss functions, and performance metrics.
• Feature Engineering: Creating and analyzing segment-based features for numeric data.
• Visualization: Plotting training metrics, confusion matrices, and feature importances to interpret model performance.
• Reproducibility: Using notebooks and Python scripts to ensure experiments can be replicated.

Repository Structure

• *.ipynb: Jupyter notebooks containing data exploration, model implementation, and experiment results.
• *.py: Python scripts for core logic, utility functions, and model definitions.
• README.md: Project overview and instructions (this file).
• requirements.txt or environment.yml: (If present) Lists required dependencies.

How to Run

Prerequisites

• Python 3.7+
• Recommended: Create a virtual environment

Installation

1. Clone the repository:

   git clone https://github.com/cupcakequeen77777/Segment-Based-Numeric-Neural-Network-Classifier.git
   cd Segment-Based-Numeric-Neural-Network-Classifier

2. Install dependencies: If you have a requirements file:

   pip install -r requirements.txt

   Or manually install core packages:

   pip install numpy pandas matplotlib scikit-learn tensorflow jupyter

Running the Notebooks

1. Start Jupyter Notebook:

   jupyter notebook

2. Open the desired .ipynb notebook and follow the cells in order. Notebooks include explanations, code, and results.

Running Python Scripts

If you prefer running scripts directly:

python your_script.py

Replace your_script.py with the desired script filename.

Results

• Trained model performance metrics (accuracy, loss, confusion matrix, etc.) are displayed in the notebooks.
• Visualizations provide insight into model learning and feature importance.